Growth, feed utilization, health and biometric parameters in Atlantic salmon (Salmo salar L.) - Influence of dietary protein-to-lipid ratio and body fat status

(1)

Philosophiae Doctor (PhD) Thesis 2018:103

Jens-Erik Dessen

Growth, feed utilization, health and biometric

parameters in Atlantic salmon (Salmo salar L.) - Influence of dietary protein-to-lipid ratio and body fat status

Vekst, fôrutnyttelse, helse og biometriske parameter hos Atlantisk laks (Salmo salar L.) – Effekt av protein-til-fett forholdet i fôret og fettinnholdet i laksen

Philosophiae Doctor (PhD), Thesis 2018:103Jens-Erik Dessen

Norwegian University of Life Sciences

Department of Animal and Aquacultural Sciences (IHA) Faculty of Biosciences

(2)

(3)

Growth, feed utilization, health and biometric parameters in Atlantic salmon ( Salmo salar L.) - Influence of dietary

protein-to-lipid ratio and body fat status

Vekst, fôrutnyttelse, helse og biometriske parameter hos Atlantisk laks (Salmo salar L.) – Effekt av protein-til-fett forholdet i fôret og fettinnholdet i laksen

Philosophiae Doctor (PhD) Thesis Jens-Erik Dessen

Department of Animal and Aquacultural Sciences (IHA) Faculty of Biosciences

Norwegian University of Life Sciences

Ås (2018)

Thesis number 2018:103 ISSN 1894-6402 ISBN 978-82-575-1569-0

(4)

(5)

Acknowledgements

First, I would like to say how fortunate and pleased I am to have obtained the possibility to take a PhD within the interesting field of aquaculture. I have learned so much during this process, both on a personal and professional plan.

First and foremost, I want to thank my main supervisor Professor Kjell-Arne Rørvik for giving me the opportunity to work in Nofima and involving me in projects that in the end has led to this PhD thesis. You are a true source of inspiration, motivation and I consider myself to be extremely lucky for being able to learn from you. Your passion, experience and knowledge within the field of aquaculture production biology is truly impressive and inspiring. You have a unique ability to teach and discuss biological challenges in the industry. If it had not been for your unique ability to follow me up and guide me through this PhD, I had never been where I am today. Thank you for all the help and the infinite number of travel days, samplings and hard work! I am forever grateful!

I also want to thank Professor Magny Thomassen for all the support and knowledge.

Your patience, wise words and ideas in projects, discussions and the writing process have been very important during the PhD process. You and Kjell-Arne have been excellent supervisor and I would like to thank both of you for giving me the opportunity to write both the master and PhD thesis at NMBU and Nofima. Thank also to my co- supervisors, Professor Turid Mørkøre, Dr. Bjarne Hatlen and Professor Bente Ruyter for your support and advice along the way. You are all very knowledgeable in your field of research and I have been lucky for being able to learn from you!

I would like to thank all co-authors for excellent collaboration and all research partners (Lerøy Midt, Blom Fiskeoppdrett, Nordlaks oppdrett and BioMar) that have made this possible. I would like to thank Jan Ivar Bildøy, Roar Paulsen, Harald Larssen, Tommy Hansen, Jostein Kjørseng, Lars Thomas Poppe, Kurt Rønning, Stig N Johnsen, Kenneth Blomvågnes, Harry Andersen, Øyvind Blom and Håvard Tennebø for great cooperation.

Thanks for all the help for technical staff at Nofima, especially thanks to Målfrid T Bjerke for the patience and helping out in the fish lab and teaching me issues related to lipid

(6)

analysis. Thanks to Sissel Nergaard for excellent assistance at Ekkilsøy. I also need to thank the people that have been involved in getting the projects that has led to this PhD- thesis. Thanks to Dr. Solveig van Nes for the effort and work in applying for R&D licenses and get them granted. Thanks to Professor Olai Einen and Dr. Hilde Toften for giving me the position as a researcher in Nofima. Thanks to Nils Haga, Dr. Arne Mikal Arnesen, Kjell-Åge Rognli, Bente Johansen, Professor Bente Torstensen and Per Brunsvik for administrative contributions and for giving us/me the opportunity to carry out these projects. Thanks to Dr. Mari Moren for your support as my group leader and for allowing me to do my PhD alongside my position as a researcher in Nofima. Thanks for involving me in projects and for all your good advice. Thanks to my fellow PhD student Rúni Weihe, for all good moments, help and good advice during the PhD process. Thank for good cooperation in connection with projects at Ekkilsøy and in the writing process.

Thanks to all past and present colleagues at Nofima for creating a super social environment. Especially I would like to thank Marta N., Sergio, Elisabeth, Trine, Torbjørn, Tove, Harald, Carlos, Carlo, John-Erik, Vibeke, Mads, Hege, Bjarne, Dimitris, Lene, Karen, Petter, Vibeke, Sissel, Kasper, Siri, Gerrit, Sten, Magnus, Solveig, Sven Martin, Christian, Øivind, Inger, Matt, Aleksei, Celest and so many others (you know who you are!!). Thank to Thomas for all sampling help, advice, learning me fishlab stuff and all good moments. Thanks to Tone-Kari (TK) ! for always supporting and encouraging me to keep on, good advice, comments on manuscripts and so on (I will start exercise soon). Thanks to Esmail for all the help and positive feedback! A super-duper thanks to Marta B! Thanks for the incredible helpfulness and that you read parts of the thesis and helping me in the final stages! I highly and sincerely appreciate it! Thanks!! --

I would like to express my warmest gratitude to family and friends for always being there for me and for making it possible for me to be where, and whom I am today. My parents: Arne and Henny, my sister Julie, little niece Dea, brother-in-law Mads, grandmother Liv and to my girlfriend Marthe. Marthe: Thanks for your endless patience, love and support-! You are amazing and the best! Finally, I have to thank my grandfather Dr. Erik Dessen. You and dad inspired and supported me to choose the pathway towards biology, aquaculture and research. We miss you!

(7)

1. Abbreviations

List of the main abbreviations used throughout this work. The rest will be described in the text as they appear.

ALT: alanine aminotransferase AP: alkaline phosphatase

CMS: cardiomyopathy syndrome

DP/DE: digestible protein and digestible energy

FCRg:Biological feed conversion ratio based on carcass weight HSMI: heart- and skeletal muscle inflammation

PD: pancreas disease PRV: piscine reovirus

P/L ratio: protein-to-lipid ratio PUFA: polyunsaturated fatty acids SAV3: Salmonid alphavirus subtype 3 S0: Under year-old smolt

S1: Year-old smolt

TGC: thermal growth coefficient VSI: visceral somatic index

(10)

2. List of publications

The thesis is based on the following publications, which will be referred to in the text by their roman numerals.

Paper I: Dessen, J. E., Weihe, R., Hatlen, B., Thomassen, M. S., & Rørvik, K. A. (2017).

Different growth performance, lipid deposition, and nutrient utilization in in- season (S1) Atlantic salmon post-smolt fed isoenergetic diets differing in protein- to-lipid ratio. Aquaculture, 473, 345-354.

Paper II: Rørvik, K. A., Dessen, J. E., Åsli, M., Thomassen, M. S., Hoås, K. G., & Mørkøre, T.

(2018). Low body fat content prior to declining day length in the autumn significantly increased growth and reduced weight dispersion in farmed Atlantic Salmon Salmo salar L. Aquaculture Research, 49(5), 1944-1956.

Paper III: Dessen, J. E., Mørkøre, T., Bildøy, J. I., Johnsen, S. N., Poppe, L. T., Hatlen, B., Thomassen, M. S., & Rørvik, K. A. (2018). Increased dietary protein-to-lipid ratio improves survival during naturally occurring pancreas disease in Atlantic salmon, Salmo salar L. Journal of fish diseases.

Paper IV: Dessen, J. E., Østbye, T. K., Ruyter, B., Bou, M., Thomassen, M. S., & Rørvik, K. A.

Sudden increased mortality in large seemingly healthy farmed Atlantic salmon (Salmo Salar L.) was associated with environmental and dietary changes.

Manuscript

(11)

3. Summary

Today, commercial diets for large farmed Atlantic salmon (Salmo salar L.) commonly contain 30-35 % protein and 35-39 % lipids, i.e. a ratio of protein-to-lipids below 1. Such energy dense diets have generally been shown to improve feed utilization and growth.

However, reducing the dietary protein-to-lipid ratio may lead to increased deposition of fat in the muscle and visceral cavity. There is evidence that high levels of body lipids may reduce feed intake and growth in salmonids, and is often referred to as lipostatic regulation. Thus, there is a risk of lowered growth prior to and during periods with increased feed intake and high fat accumulation when feeding high-fat diets. The present thesis test the hypothesis that increased dietary protein-to-lipid ratio, and the possible involvement of lipostatic regulation on body fat levels can be utilized to significantly improve fitness-related traits including growth, survival and nutrient deposition of farmed Atlantic salmon.

Paper I describes the effects of isoenergetic diets with different protein-to-lipid (P/L) ratio on growth, feed intake, feed conversion, biometrics, nutrient retention and deposition in S1 Atlantic salmon post-smolt. The study was conducted during the early seawater phase from April to September. Significantly lower muscle fat, whole body lipid, and energy level was observed in the post-smolt fed high compared to low P/L ratio in July, approximately three months after the trial started. Reduced level of muscle fat/body fat in July significantly improved feed intake, growth and weight gain compared to fish fed low P/L ratio from July to September. The high P/L ratio group had also a significantly lower feed conversion ratio based on gutted weight (FCRg). In line with this, the visceral somatic index (VSI) of the group fed the high dietary P/L ratio was relatively stable during the experiment, whereas the VSI of the group fed the low dietary P/L ratio increased gradually, resulting in significantly higher VSI at the end of the trial.

The increased protein content in the high P/L ratio diets was efficiently utilized for growth and weight gain, assessed by nutrient retention, particularly in the last period of the trial.

(12)

In Paper II, low and high P/L ratios and restricted ration (~ 50%) of the high P/L diet were used to alter the lipid deposition prior to the autumn in large salmon. In this study, a clear treatment effect on body fat was observed after three months (May - August).

The salmon fed low P/L ratio had higher fat content than those fed high P/L ratio, both in muscle (16.4 vs. 13.2 %) and viscera (39 vs. 29 %). Restricting the ration of the high P/L diet to 50 % further reduced the fat content to 11.3 % in muscle and 27 % in the viscera. Tagged individuals from the groups with different lipid content were restocked and mixed in the same pens, and then fed the same diet for seven months (August – March). As in Paper I, reducing the level of muscle fat prior to autumn significantly increased growth and weight gain from August to October. In other words, the weight gain was the highest for the restricted group, intermediate for the high P/L ratio and the lowest for the low P/L ratio group. In October, after two months of feeding a common diet, the muscle fat content was similar in fish from all three groups, whereas the differences in visceral fat content disappeared after four months (December). Although the differences in body weight, length and lipid content between the groups had been offset, the high P/L ratio and restricted group showed a significantly increased growth compared to the low P/L-ratio group in the latter stages of the trial (December – March), resulting in an overall weight gain difference of up to 1 kg. The high P/L ratio group had significantly higher final body weight, whereas the restricted group ended up with a numerically higher final body weight than in the low P/L ratio group. In addition, the variation in body weight and shape was significantly higher in the low P/L ratio group.

The results from Paper I and II demonstrate that in early autumn, the salmon is able to replenish lipid stores rapidly after dietary lipid restriction and that energy intake and storage is of high priority. In Paper I, unlike Paper II, low and high P/L ratio were fed throughout the trial and it is therefore not possible to isolate the direct effect of diet on growth from the indirect effect caused by different body fat accumulation. Hence, paper I may also indicate that dietary P/L-ratio of 1.12 (DP/DE of 15.2 g MJ kg^-1) was below the P/L ratio for optimal growth during the early seawater phase for S1 salmon. This result is in line with previous studies, although now verified using isoenergetic diets under ambient environmental conditions.

(13)

Paper III describes the effects of increased dietary P/L ratio for S0 salmon on mortality rates, biometric and quality related parameters during the entire grow-out period in sea, within the SAV3 endemic zone. The low P/L group was fed a typical standard diet with 35% protein and 35% fat (P/L: 1), versus the high P/L group that was fed a diet with 47% protein and 24% fat content (P/L: 2). During the first summer at sea, a co-infection of SAV3 and PRV was detected and a natural PD outbreak was observed. The increased dietary P/L ratio improved survival during the natural outbreak of PD. In addition to diet, body weight and delousing mortality (induced stress) prior to the PD outbreak were also found to contribute significantly to explain the observed variation in PD associated mortality. The high P/L group had a mean mortality rate of 1.9 %, whereas the low P/L group had a mean mortality rate of 3.7 %. Subsequent to the PD outbreak, a large amount of fish failed to grow and caused an accumulation of runts (severely thin diseased fish). At the end of the trial, a significantly lower amount of runt fish was detected among fish fed high P/L ratio (12 vs. 21%) and among large compared to small body weight groups (11 vs. 20%) prior to PD.

In Paper IV, an event of sudden mortality of large (2.5 kg) seemingly healthy farmed salmon during the winter period in northern Norway is reported. The experimental fish were reared in four net-pens and two dietary treatments were established; a high or low P/L ratio diets. An increased mortality (of 6 and 10%) was only observed within the two net-pens receiving the high P/L ratio experimental diets, following an abrupt reduction in dietary P/L ratio (increase in dietary fat level) six weeks earlier. The moribund/dying fish had significantly higher lipid content in the liver, altered hepatic fatty acid composition, and increased levels of ALT and AP in the blood plasma compared to non- dying fish, indicating impaired hepatic function. A possible hypothesis involving reduced recruitment of fat cells in high P/L-salmon is presented.

Taken together, the results from this thesis show that alterations in dietary protein-to- lipid ratio have profound potential effects on growth, lipid deposition, nutrient retention and health of farmed Atlantic salmon. . The results obtained during this work related to fat deposition and subsequent growth may be crucial knowledge when developing new dietary concepts in semi-closed and closed RAS production units, where water temperature and photoperiod can be manipulated.

(14)

Sammendrag

I dag inneholder kommersielle dietter for stor oppdrettet Atlantisk laks (Salmo salar L.) vanligvis 30-35% protein og 35-39% fett, dvs. et forhold mellom protein-til-lipider under 1. Slike energitette dietter har generelt vist seg å kunne forbedre fôrutnyttelse og vekst. Imidlertid kan en reduksjon av protein-til-lipid forholdet i fôret føre til økt deponering av fett i muskel og rundt innvollene. Det er vist at høye nivåer av kroppsfett kan redusere fôrinntak og vekst hos laksefisk og blir ofte referert til som lipostatisk regulering. Det er derfor en økt risiko for redusert vekst dersom man benytter et fôr med et høyt fettinnhold før og under perioder hvor fôrinntaket og fettakkumulering er høy. Denne avhandlingen tester hypotesen om at økt protein-til-lipid-forhold i fôret til laks, samt en mulig involvering av lipostatisk regulering ved å redusere kroppsfettet, kan utnyttes for å forbedre egenskaper som tilvekst, overlevelse og fôrutnyttelse hos oppdrettslaks.

Artikkel I beskriver effekten av isoenergetiske dietter med forskjellig protein-til-lipid (P/L) forhold på vekst, fôrinntak, fôrutnyttelse, biometri og retensjon av næringsstoffer hos S1 post-smolt. Denne studien ble gjennomført fra sjø-utsett i april til september. Det ble funnet et signifikant lavere nivå av fett og energi i muskel og helkropp for laks gitt et fôr med høyt sammenlignet med lavt P/L forhold i juli, omtrent tre måneder etter at forsøket startet. Det redusert nivå av kroppsfett i juli forbedret fôrinntaket, veksten og vektøkningen betydelig sammenlignet med fisken som ble gitt et lavere P/L forhold i perioden juli til september. Gruppen gitt et høyt P/L forhold hadde også bedre fôrutnyttelse basert på sløyd vekt (FCRg). I tråd med dette var den relative innvollsvekten (viscerale somatiske indeksen, VSI) for gruppen gitt et høye P/L forholdet relativt stabilt under forsøket, mens VSI for gruppen gitt et lavere P/L forhold økte gradvis, noe som resulterte i signifikant høyere VSI på slutten av forsøket. Det økte proteininnholdet i diettene med høyt P/L forhold ble effektivt utnyttet for vekst og vektøkning, vurdert ved retensjonsberegninger, særlig i den siste perioden av forsøket.

(15)

I artikkel II, ble lave og høye nivåer P/L forhold i fôret, samt begrenset rasjon (~ 50%) av det høye P/L fôret benyttet for å endre fettnivået/status før høsten i stor laks. I denne studien ble det påvist en klar effekt av de ulike fôrbehandlingen tre måneder etter forsøksstart (mai - august). Gruppen gitt et lavt P/L forhold hadde høyere fettinnhold enn gruppen gitt et høyt P/L forhold, både i muskel (16,4 vs. 13,2%) og innvollsmassen (39 vs. 29%). Ved å gi halv rasjon av fôret med et høyt P/L forhold ble fettet deponering reduserte ytterligere til 11,3% i muskel og 27% i innvollsmassen. Gruppene (markert med PIG-tagg) med forskjellig fett innhold ble deretter overført til samme enheter/merder oppsamlet og gitt lik diett (samme P/L forhold) i syv måneder (august - mars). I likhet med artikkel 1 ble det vist at gruppene med et redusert nivå av kroppsfett før høsten hadde økt tilvekst fra august til oktober. Vektøkningen var høyere for gruppene gitt halv rasjon og et høyt P/L forhold sammenliknet med gruppen gitt et fôr med lavt P/L forhold. I oktober, etter to måneder med fôring av lik diett, var det ingen forskjeller i muskelfett mellom de ulike gruppene og forskjellene innvollsfettet var borte etter fire måneder (desember). Selv om forskjellene i kroppsvekt, lengde og fettinnhold mellom gruppene var blitt kompensert, ble det observert en signifikant økning i tilvekst for gruppen gitt det høye P/L forholdet og halv rasjon av dette fôret (fra mai-august) sammenlignet med gruppen gitt et lave P/L forhold i den siste perioden av forsøket (desember - mars). Dette resulterte i en total relativt vektøkning for gruppene gitt halv rasjon og et høyt P/L forhold på opptil 1 kg. Gruppen gitt et høyt P/L forhold hadde derfor en signifikant høyere sluttvekt, mens gruppen gitt halvrasjon av dette fôret endte opp med en numerisk høyere sluttvekt sammenliknet med gruppen gitt et lave P/L forhold. I tillegg var variasjonen i fiskevekt og kroppsform signifikant høyere i gruppen gitt lavt P/L forhold.

I artikkel III blir effekten av et økt P/L forholdet i fôret for S0 laks på overlevelse, biometriske registeringer og kvalitets parametere i løpet av hele sjøfasen, innenfor den SAV3-endemiske sonen testet. En gruppe ble fôret med et standard høyfett-fôr med 35%

protein og 35% fett (P/L: 1), men den andre gruppen ble gitt et fôr med 47% protein og 24% fett (P/L: 2). I løpet av den første sommeren i sjø ble det oppdaget en samtidig infeksjon av SAV3 og PRV, og det ble observert et naturlig utbrudd av PD. Økt forhold mellom P/L i fôret forbedret overlevelsen under det naturlige utbruddet av PD. I tillegg

(16)

til diett ble kroppsvekt og dødelighet ved avlusing (indusert stress) før PD utbruddet også funnet å bidra betydelig til å forklare den observerte variasjonen i PD-relatert dødelighet. Gruppen gitt et fôr med høyt P/L forhold hadde en gjennomsnittlig dødelighet på 1,9%, mens gruppen gitt lavt P/L forhold hadde en gjennomsnittlig dødelighet på 3,7%. Etter PD-utbruddet mislyktes en stor mengde fisk med å gjenoppta inntak av mat og forårsaket en kraftig akkumulering av såkalte «runts/taperfisk», som er alvorlig avmagret klinisk syk fisk. På slutten av forsøket ble det registrert en signifikant lavere mengde med runts blant gruppen gitt et fôr med høyt P/L forhold (12 vs. 21%) og blant stor sammenliknet med liten kroppsvekt (11 vs. 20%) før PD utbruddet inntraff (1.9 vs. 1.3 kg).

Artikkel IV beskriver en hendelse med plutselig økt dødelighet av stor (2,5 kg) tilsynelatende frisk oppdrettslaks i løpet av vinteren i Nord-Norge. I dette forsøket ble fisken oppdrettet i fire merder, hvorav to ble gitt en fôrserie med høyt P/L forhold og de to andre et fôrserie med lavt P/L forhold. Den økte dødelighet (på 6 og 10%) ble kun observert i de to enhetene/merdene som ble gitt fôrserien med høyt P/L forhold først etter at fettinnholdet i denne fôrserien ble økt (reduserte P/L forholdet) seks uker tidligere. Den døende fisken hadde betydelig høyere fettinnhold i leveren, endret fettsyresammensetning, og økte nivåer av ALT og AP i blodplasma sammenlignet med ikke-døende frisk fisken. Dette kan indikere en nedsatt leverfunksjon hos den døende fisken, som trolig har negative konsekvenser for helse og robustheten til denne fisken.

En mulig hypotese som involverer redusert rekruttering av fettceller i laks gitt et høy P/L forhold blir presentert.

Samlet sett viser resultatene fra denne avhandlingen at endringer i protein-til-lipid forhold i fôr til oppdrettslaks har store effekter på fôrinntak, tilvekst, fett deponering og helse. potensielle effekter på vekst, lipidavsetning, næringsreservering og helse av oppdrettslaksatlantisk laks. Resultatene i denne avhandlingen knyttet til fettdeponering og påfølgende vekst, kan være viktig kunnskap ved utvikling av nye fôrkonsepter i semi- lukkede og lukkede RAS-produksjonsenheter hvor vanntemperatur og fotoperiode kan manipuleres.

(17)

4. General introduction 4.1 Atlantic salmon aquaculture

In the world aquaculture production of diadromous fish, Atlantic Salmon (Salmo salar L.) is the dominating fish species, with a total production of 2.38 million tons in 2015 (FAO). Norway is the world leading producer and exporter of Atlantic salmon, with a sale of over 1.23 million metric tons, representing a first-hand value of 61.6 billion NOK in 2017 (Statistics Norway, 2018; figure 1). Atlantic salmon was first cultivated in the beginning of the 1970s in Norway, and the industry began with small family-owned businesses. Today, the farming of Atlantic salmon is a modern, intensive and globalised industry, with large multinational companies involved in rearing, feed production and processing of farmed salmon. The exponential increase in production can be attributed several factors, such as genetic selection, research & development, technical innovations, improved inputs, and production practices. This tremendous productivity growth has allowed production cost to be reduced, making farmed salmon a high quality competitive product (Asche, 2008).

Figure 1 Sales of Atlantic salmon (metric tons) and first hand value (NOK million) from 1992 to 2017. Source: Statistics Norway, aquaculture (2018) (“Statistics Norway,” 2018)

(18)

4.2 Feed for farmed Atlantic salmon

The traditional salmon diets in 1990s contained up to 90 % of marine ingredients, mainly consisting of fish meal (FM) and fish oil (FO) (Ytrestøyl et al., 2015). Current modern diets contain less than 30% of marine ingredients, and this has been due to a static production of FM and FO during the last decades, combined with the fact that wild fisheries are being fully- or over-exploited (FAO, 2012; Ytrestøyl et al., 2015). In addition, the competitive pressure for FM and FO is high, due to that these sources are also used in feed production for other farmed livestock and in pharmaceutical industries. To reduce the dependency on wild fisheries and maintain an increase in sustainable aquaculture production, protein and oil sources of vegetable origin have been introduced and are used together with marine ingredients in modern diets for salmon (Turchini et al., 2009; Ytrestøyl et al., 2015).

In current intensive aquaculture production, high energy extruded diets are extensively used for Atlantic salmon. The introduction of extrusion technology together with vacuum coating has led to an increase of the dietary lipid level in salmonid diets during the last decade. Traditional pelleted diets used from the 1970s until the 1990s had a protein content of 45-55 % and lipid content of 10-20 %. Today, extruded vacuum coated diets containing 30-35 % protein and 35-39 % lipids are commonly used as grow-out diets (Torrissen et al., 2011). In Norway, the increase in dietary lipid inclusion has partly been driven by governmental restrictions (feed quotas), in order to control the biomass and restrict production growth from 1996 until 2005. The response from the aquaculture industry to this restriction was to increase the dietary lipid level (increasing the dietary energy) in order to improve feed utilization, and thereby gain higher fish biomass with less feed inputs (Torrissen et al., 2011). Protein sources are a major input cost in aquaculture feeds, and the global demand and prices for protein sources have been and are high. Lipids are generally a cheap source of energy compared to protein, and overall reductions in dietary protein together with an increase in lipid level are often utilized to reduce feed prices and increase sustainability. The proportion of lipid in salmon diets has approximately doubled during the last 30-40 years (Torrissen et al., 2011).

(19)

Today there is a variety in dietary protein-to-lipid ratio of commercial salmon feeds.

However, the general trend is that the lipid content is gradually increased with higher fish weights, regardless of season. Feed costs account for more than 50 % of the total production costs in salmon farming, and fish farmers perception of low feed prices acts partially as a driver of the feed industries focus on producing the cheapest possible feed.

Salmonids have high ability to utilize large amounts of lipids in high-energy diets efficiently for growth, resulting in good feed conversion and a favourable protein sparing effect (Azevedo et al., 2004; Karalazos et al., 2011, 2007). Based on these factors and the high demand and prices of preferred protein sources, it is likely to assume that the trend of high dietary lipid content can affect the protein-to-lipid ratio in diets during the whole seawater phase. However, small post-smolt that undergo rapid body growth require a high portion of digestible protein than larger salmon (Einen and Roem, 1997).

It is therefore important that the protein content in current and future dietary regimes support optimal growth and health at different fish sizes and phases of the seawater production.

4.3 Diseases in salmon aquaculture

A major challenge during farming of Atlantic salmon is high levels of mortalities observed during the seawater phase. In Norwegian salmon aquaculture, the total mean production losses during the seawater phase was 13.1 % from 2010 to 2017 (“Statistics Norway,” 2018). Large proportions of this loss are due to mortality caused by viral diseases (Hjeltnes et al., 2018). Currently, the most widespread, severe fatal viral diseases in Norwegian aquaculture are pancreas disease (PD), heart- and skeletal muscle inflammation (HSMI), and cardiomyopathy syndrome (CMS) (Hjeltnes et al., 2018, figure 2).

(20)

Figure 2 Number of Norwegian production sites diagnosed with the viral diseases pancreas disease (PD), heart- and skeletal muscle inflammation (HSMI), and Cardiomyopathy syndrome (CMS) from 2007 – 2017. Data sources: Hjeltnes et al. (2018).

4.3.1Pancreas disease

PD is a widespread contagious viral disease affecting Atlantic salmon and rainbow trout (Oncorhynchus mykiss) and is a significant problem in the European salmonid farming industry (Graham et al., 2011; Jansen et al., 2017). The first recognized and described cause of the disease was conducted in Scotland in 1976 (Munro et al., 1984), whereas the first detected cases in North America and Norway were in 1987 and 1989, respectively (Kent and Elston, 1987; Poppe et al., 1989). Salmonid alphavirus (SAV) is the causative agent of PD in farmed Atlantic salmon and rainbow trout, and is allocated to the genus alphavirus within the family Togaviridae (Weston et al., 1999). In Norway, the SAV subtype 3 and a marine subtype 2 have been detected (Hjortaas et al., 2013;

Hodneland et al., 2005). Since 2003, PD caused by SAV3 has been endemic along the west coast of Norway up to Hustadvika in Møre and Romsdal (63° latitude – SAV3 endemic zone), particularly in counties of Hordaland and Rogaland (Jansen et al., 2015;

Jensen and Gjevre, 2018). PD can occur throughout the yearly cycle, but the risk of outbreaks is highest during the spring and summer months when the seawater

(21)

temperature is increased (Hjeltnes et al., 2017; McLoughlin and Graham, 2007; Rodger and Mitchell, 2007; Stene et al., 2014). PD causes pathological changes that involve partly or severe loss of exocrine pancreatic tissue, cardiac and skeletal myopathies, epicarditis and white skeletal muscle degeneration and/or inflammation (Christie et al., 2007; McLoughlin and Graham, 2007; Taksdal et al., 2007). Mortality rates can reach up to 63% for sites that are severely affected by PD and among surviving fish, subsequent failure to grow may cause thin fish with poor condition (runts) and high number of discarded fish at slaughter (reviewed by Jansen et al., 2017). Several studies have found that PD may also impair the fillet quality of slaughter sized salmon (Larsson et al., 2012;

Lerfall et al., 2012; Taksdal et al., 2012). A newly published economical simulation showed that PD caused a total direct cost for Norwegian fish farmers in 2015 of 2.4-2.8 billion NOK (Vedeler, 2017). This was equivalent to an increase in production cost of about 2.2 NOK/kg head on gutted salmon (HOG).

4.3.2Heart and skeletal muscle inflammation

HSMI is another common fatal disease of farmed salmon and the disease have been link to the piscine orthoreovirus (PRV) (Palacios et al., 2010). However, PRV seem to be ubiquitous among farmed salmon in Norway (Løvoll et al., 2012), and can be present in high titers without causing mortality or marked lesion in the heart (Garseth et al., 2013).

Severe HMSI outbreaks gives direct inflammatory lesions in cardiac and skeletal muscle and such damage may occur at an early stage in the disease progression, and may persist for many months after clinical disease (Kongtorp et al., 2006). The lesions observed during outbreaks of HSMI are similar to those described for PD and CMS (R. T. Kongtorp et al., 2004). The histopathological changes associated with HSMI in salmon has previously been thoroughly described, of which epi-, endo- and myocarditis, myocardial and skeletal muscle necrosis and signs of liver damage are central features (R T Kongtorp et al., 2004). Outbreaks of HSMI may lead to lowered appetite, reduced feed utilization and increased mortality, and although the mortality and duration of the outbreak can vary, mortality rates up to 20 % are observed (Alne et al., 2009; Kongtorp et al., 2006). Natural outbreaks of HSMI have normally been recorded 5 to 9 months after transfer to sea. However, observation shows that outbreaks may occur during the

(22)

whole seawater phase and as early as 14 days following seawater transfer (Bornø and Lie, 2015; R. T. Kongtorp et al., 2004).

4.3.3Cardiomyopathy syndrome and non-infectious cardiovascular disease

CMS is a serious cardiac related disease that affect large Atlantic salmon in sea and the diseases is associated with the totivirus Piscine myocarditis virus (PMVC) (Hjeltnes et al., 2018). Salmon that are affected by CMS are presumably in good condition prior to the time of death, which occur suddenly (Brun et al., 2003). CMS is regarded as a chronic disease, which can cause prolonged periods of moderate to elevated mortality (Brun et al., 2003). Mortality of seemingly healthy salmon has also been linked to non-infectious cardiovascular diseases/failures and are often observed in salmonid aquaculture (Dalum et al., 2017; Hjeltnes et al., 2018; Poppe et al., 2007; Tørud and Hillestad, 2004).

Stress is generally related to the outbreaks of diseases and increased mortality caused by HSMI, PD and CMS are often reported in association with handling and operation measures at site level, e.g. delousing and relocating fish (Bornø and Lie, 2015; Hjeltnes et al., 2018).

4.3.4Disease prevention

In terms of viral disease prevention, a vaccine against PD is available in Norway since 2007 and recently a DNA-vaccine has been introduced. However, the immunity of the DNA-vaccine (Clynav, Elanco Europe Ltd) has only shown a duration of three months post vaccination (Felleskatalogen, 2018). In a cohort study conducted by Jensen et al.

(2012), PD vaccinated fish had lower PD-associated mortality, number of discarded fish at slaughter and better growth than non-vaccinated fish. However, the efficacy of the monovalent vaccines under field conditions has been questioned and high PD associated mortality has been observed in PD vaccinated fish (Hjeltnes et al., 2018). No commercially available vaccines for HSMI and CMS seem to exist. However, several breeding companies are now marketing genetic lines with increased resistance towards PD, HSMB and CMS, using quantitative trait locus (QLT) analysis and marker-assisted selection.

(23)

Today, there are several commercially available feeds designed for better performance in connection with viral infections, for clinical use after infection or as feeds with prophylactic, immune stimulating or anti-inflammatory effects (often referred to as functional feeds). The effectivity of some of these functional feeds has been reported in scientific studies. Low-lipid diets containing high levels of specific polyunsaturated fatty acids (PUFAs) have been used as a potent tool to increase the tolerance/resistance towards HSMI and CMS in Atlantic salmon by modulating tissue fatty acid composition and eicosanoid production (Martinez-Rubio et al., 2014, 2012). Commercial available

“PD feeds” are frequently used within the SAV3 endemic zone (Jansen et al., 2015), and these feeds are often formulated to contain lower amounts of lipids and increased levels of protein (pers. comm. 2016). However, no scientific studies are published on the potential effects of such feeds related to PD associated mortality of large salmon. High intake of dietary lipids is associated with metabolic risk factors (Johnson et al., 2008;

Nicholls et al., 2006), and it can be suggested that several diseases in salmon aquaculture are worsened by the use of high-fat diets. This can be particularly prominent for diseases that also affect the heart, such as PD, HSMI and CMS.

4.4 Growth and feed utilization

Attaining optimal growth is one of the most important parameter in aquaculture, which defines the production efficiency and is a good indicator of a robust and healthy fish. The growth rates of salmonids depend on feed intake and utilization, which are highly dependent on water temperature, photoperiod, and affected a wide range of other internal and external factors such as genetics, health status, adiposity, physiological processes (smoltification, maturation etc.), water quality, fish size, dietary composition and feeding regime (Aksnes et al., 1986; Austreng et al., 1987; Bendiksen et al., 2003;

Einen and Roem, 1997; Gjedrem, 2000; Sveier and Lied, 1998; Thorarensen and Farrell, 2011). The thermal growth coefficient (TGC) is a highly used growth model for fish, that express growth rate independent of temperature and fish size, which makes the model both useful and flexible (Cho, 1992; Thorarensen and Farrell, 2011). As other models, the TGC has some assumptions and limitations (Jobling, 2003), but is relatively robust within the normal temperature range (Thorarensen and Farrell, 2011). Feed conversion

(24)

ratio (FCR) is a useful indicator of feed utilization and efficiency, and describes how much feed is required to produce 1 kg of fish (direct inputs to outputs). Both the TGC and FCR vary to a large degree between different growth related studies of Atlantic salmon conducted both in net-pens and tanks, and an excellent overview of this is given in Thorarensen and Farrell (2011), showing TGC from 0.1 to 4.8 and FCR from 0.7 to 1.7.

As discussed in this article, these differences show that salmon growth rates depend on experimental conditions, and they reflect the seasonal cycles in growth performance of Atlantic salmon. High growth rates and efficient production during the seawater phase are fundamental goals for all fish farmers. Thus, factors that can limit and increase salmon growth are of great importance.

4.5 Seasonal variations in growth and lipid deposition

The majority of farmed Atlantic salmon is reared in net-pens that are exposed to seasonal environmental changes. Several studies report seasonal variation in growth and lipid deposition of Atlantic salmon, mainly due to changes in seawater temperature and photoperiod (Alne et al., 2011; Mørkøre and Rørvik, 2001; Nordgarden et al., 2005, 2003c, 2003a; Weihe et al., 2018). However, some of these variations also seem to be influenced by smolt-type and timing of sea transfer. In the salmon industry, smolt are regularly transferred to the sea in the spring as “in season” year-old smolt (S1) or in the autumn as “out of season” under-year-old smolt (S0). Seawater adaption is an energy demanding process, and a reduction in appetite, growth and body lipid content is often observed after sea transfer, particularly for S1 post-smolt during the spring (Alne et al., 2011; Toften et al., 2003; Usher et al., 1991). S0 smolt on the other hand, are in some studies reported to not experience low-performing periods after sea transfer in autumn (Alne et al., 2011; Lysfjord et al., 2004). In line with this, the S1 salmon show increased feed intake, growth and lipid deposition during the late summer and autumn period (Alne et al., 2011; Mørkøre and Rørvik, 2001). However, it is shown that S0 smolt reared in mid-west part of Norway can experience reduced feed intake, growth and lipid deposition during the first spring in sea, 5-7 months after sea transfer (Alne et al., 2011).

High feed intake and TGCs are generally observed summer and until late autumn for large salmon (Mørkøre and Rørvik, 2001; Nordgarden et al., 2003a), whereas reduced

(25)

feed intake, TGC and FCR are often observed during the winter period (Mørkøre and Rørvik, 2001). Photoperiodic manipulation using continuous light during the winter and early spring period is shown to induce increased feed intake and growth during the spring and early summer (Nordgarden et al., 2003a). Taken together, seasonal shift with its associated environmental conditions and internal factors seem to induce metabolic changes that significantly affect growth and feed utilization in salmon.

Important factors as body size, dietary lipid content and feed ration are shown to alter lipid deposition and fat content in salmonids (Hemre and Sandnes, 1999; Hillestad et al., 1998; Shearer, 1994; Torstensen et al., 2001). In relation to season, farmed salmon show high fat deposition and increased condition factor, with a concomitant increase in feed intake and weight gain during the summer and autumn period (Mørkøre and Rørvik, 2001; Nordgarden et al., 2003b). This pattern is particularly pronounced for S1 salmon at high latitudes that experience long winter and late spring (Mørkøre and Rørvik, 2001). Prolonged high seawater temperatures and declining day length characterizes the autumn period. During the late autumn and winter, a decline or stagnation in the level of muscle fat in salmon is often observed, which relates to reduced feed intake and increased FCR during low sea temperatures and short day length (Mørkøre and Rørvik, 2001). This pattern of fat deposition is commonly observed in commercial farming of salmon and during large-scale studies under ambient environmental conditions.

4.6 Fat storage as a reproductive adaption?

A model develop by Professor Kjell-Arne Rørvik links the pattern of fat storage to environmental adaption and the reproduction life-strategy of Atlantic salmon. Sexual mature Atlantic salmon spawn during the late autumn period. The sexual maturation process requires, in addition to photoperiodic stimuli, sufficient fat and energy reserves (Kadri et al., 1997, 1996; Rowe and Thorpe, 1990; Taranger et al., 2010). Development of gonads are energetically demanding process and require severe energy investment (Fleming, 1996; Jonsson et al., 1997). Appropriate and available fat reserves during the spring period seems to be a major factor controlling the initiation and progression of the maturation process in salmon (Thorpe, 1994; Thorpe et al., 1998; Wright, 2007). To low energy and fat levels may arrest the maturation process and postpone reproduction

(26)

(Duston and Saunders, 1999; Rowe et al., 1991; Rowe and Thorpe, 1990; Thorpe, 1994;

Thorpe et al., 1990). To assure sufficient energy stores, salmon need to utilize the late summer and early autumn for accumulation of fat and build-up of energy stores prior to the initiation of the maturation process the following spring. This reproductive fat storage cycle hypothesis is illustrated in figure 3. The maturation process starts and continues if the salmon has sufficient energy stores in the spring before (black arrow).

Thus, the overall reproduction process starts with the crucial accumulation of fat and build-up of energy stores during late summer and early autumn prior the initiation of the maturation process the following spring (grey arrow), one year ahead of spawning.

The fat content in the beginning of the autumn may therefore be a dominant factor for feed intake, growth and the accumulation of fat.

Figure 3 The fat storage cycle hypothesis. Atlantic salmon spawn during the autumn and the maturation process starts (in addition to light stimuli) and continues if the salmon has sufficient energy stores in the spring prior to the autumn (black arrow). The overall reproduction process start with the accumulation and build-up of the energy stores during late summer and autumn period prior to the initiation of the maturation process the following spring (grey arrow), one year ahead of spawning. The fat content prior to the autumn may therefore be of importance.

(27)

4.7 Compensatory growth and lipostatic regulation

Healthy animals exposed to optimal environmental and nutritional conditions display good growth, whereas animals that encounter setbacks induced by nutritional deficits or sub-optimal conditions often display accelerated growth rate to recover lost body mass when circumstances are normalized (Ali et al., 2003; Arendt, 1997; Metcalfe and Monaghan, 2001). The majority of this evidence arise from studies in which animals exhibit accelerated growth after a period of growth depression, often referred to as compensatory growth (Ali et al., 2003; Dobson and Holmes, 1984; Hayward et al., 1997;

Jobling, 2010). The compensatory growth response phenomena has been observed in several animals, including several fish species (Ali et al., 2003; Jobling et al., 1993;

Sæther and Jobling, 1999; Wilson and Osbourn, 1959). The most common method to provoke compensatory growth responses in fish is by the means of complete or partial food deprivation prior to periods with food availability (Ali et al., 2003; Dobson and Holmes, 1984; Nikki et al., 2004). However, compensatory growth may also occur for fish periodically exposed to low temperatures, hypoxia and disease treatment (Foss and Imsland, 2002; Mortensen and Damsgård, 1993; Speare and Arsenault, 1997). The degree of compensatory growth in fish vary and is often categorized based on the capacity of the fish to catch-up, and hence to achieve lower, similar or higher size/mass as their non-restricted counterparts, referred to as partial-, complete- and over compensation, respectively (Ali et al., 2003).

Feed restriction or deprivation induce changes in body energy by depleting lipid stores, and during the course of compensatory growth and hyperphagia, body weight and lipid reserves are gradually restored (Ali et al., 2003; Bull and Metcalfe, 1997; Jobling and Miglavs, 1993; Metcalfe and Thorpe, 1992). The lipostatic model is often discussed within the circumstances of compensatory growth responses in fish (Jobling and Johansen, 1999; Johansen et al., 2001), and this identifies adipose tissue and stored lipids as important factors governing appetite (Jobling and Johansen, 1999; Keesey and Corbett, 1984; Kennedy, 1953). The model implies that the amount of stored fat has a negative feedback control on feed intake and is important for the regulation of energy

(28)

homeostasis. Hence, compensatory growth is not only a response to recover body weight, but also a strong response to restore lipid levels and compensatory growth will therefore cease once this is achieved (Ali et al., 2003; Jobling and Johansen, 1999;

Johansen et al., 2002). Johansen et al. (2002) showed that altering body lipids of juvenile salmon by feeding low-fat diets induced similar growth responses as deprivation and feed restriction.

Salmonids increase the deposition of fat in the muscle and visceral cavity as the fat content in the feed increases (Bendiksen et al., 2003; Hillestad et al., 1998). In view of the possible involvment of lipistatic regulation, it can be assumed that a diet with a lowered lipid level but with sufficient energy content (increasing the dietary protein-to- lipid ratio), can be an approach to reduce the deposition of lipids and enhance feed intake. It may also indicate that lowering the body fat levels may increase feed intake and growth, particlarly prior to or during periods with high fat acummulation.

.

(29)

5. Objectives

The general objective of the present thesis was, in both small- and large-scale studies, to investigate the impact of dynamic changed dietary protein-to-lipid ratios throughout the production cycle of farmed Atlantic salmon with focus on:

x The effects of isoenergetic diets with different protein-to-lipid ratio on growth, feed intake, feed conversion, biometrics, nutrient retention and deposition in S1 Atlantic salmon post-smolt (Paper I).

x The influence of body fat levels prior to the autumn on subsequent growth, weight dispersion, biometrics and lipid deposition in Atlantic salmon. (Paper II).

x The effects of seasonal changes in dietary protein-to-lipid ratio on growth, survival and quality in Atlantic salmon during health challenging periods under commercial condition in the southern (Paper III) and northern (Paper IV) Norway.

Based on these aims, the initial main hypothesis was that increased dietary protein-to- lipid ratio and the possible involvement of lipostatic regulation on body fat levels can be utilized to significantly improve fitness-related traits including growth, survival and nutrient deposition of farmed Atlantic salmon.

(30)

24

e nt al o v er v ie w

(31)

7. Main results and general discussion

The following section summarizes and discusses the main results from Paper I–IV. In this section, protein-to-lipid ratio is referred to as P/L ratio. To simplify the discussion, the term high and low dietary P/L ratio is used instead of the specific P/L ratio for each experimental period within Paper I-IV. Table 1 summarizes the specific P/L ratios that were used in Paper I-IV and how they are referred to in the text (high or low). The seasonal timing of the dietary induced changes from low to high lipid level within the low P/L groups (test variable), is shown in figure 4.

Table 1 Overview of the specific protein-to-lipid (P/L) ratios and the corresponding ratios between digestible protein and energy (DP/DE) used for each experimental period within Paper I-IV, and how they are referred to in the text (high or low).

P/L ratio DP/DE ratio

Paper Time period High Low High Low

I April - June 1.86 1.55 20.5 18.5

June - July 1.59 1.26 18.9 16.5

July - September 1.40 1.12 17.4 15.2

II May - August 2.85 0.98 23.7 13.5

August - December 1.33 16.6

December - March 0.95 13.5

III During PD outbreak 2.00 1.01 21.6 15

IV May - August 1.85 1.43 -

August - October 1.64 1.31 -

October - December 1.52 1.01 -

December -

February 1.17 1.02 -

March - July 1.33 1.02 -

PD; Pancreas disease

(32)

Figure 4 The seasonal timing of the dietary induced changes from low to high lipid level within the high P/L groups (test groups).

7.1 Body fat levels and growth

For the control groups, the lipid deposition observed in Paper I, II and IV are generally in line with previous studies showing elevated lipid deposition with increased body weight, feed intake, dietary lipid content and growth (Aksnes, 1995; Einen et al., 1999; G I Hemre and Sandnes, 1999; Hillestad et al., 1998; Mørkøre and Rørvik, 2001; Shearer, 1994; Torstensen et al., 2001). The amount of stored fat seems to be an important regulator of appetite and feed consumption in fish and high fat levels correlates with subsequent reductions in feed intake in salmonids (Johansen et al., 2003; Silverstein et al., 1999). To evaluate the effect of altered lipid deposition on growth may be difficult in salmonids, since they tolerate long periods with food restriction or deprivation before

(33)

fat stores are changed. Factors like duration, fish size, dietary composition, environmental cues and life stage are important to consider in the planning of such trials and when interpreting the results. The different dietary P/L ratio used in Paper I, II and IV triggered significant effects on lipid deposition in the muscle and visceral cavity, which are the main lipid stores in salmon (Aursand et al., 1994; Sheridan, 1994). In Paper I, significantly lower muscle fat, whole body lipid, and energy level was observed in the post-smolt fed high compared to low P/L ratio in July, approximately three months after the trial started. A negative correlation between the level of muscle fat in July and the feed intake from July to September was observed. The fish fed high P/L ratio displayed a significantly higher feed intake, growth and weight gain compared to fish fed low P/L ratio from July to September. At the end of the trial, no differences in muscle fat was observed between the groups. High body lipid levels may reduce feed intake and growth in salmon (Johansen et al., 2003), and this observation may indicate a possible involvement of lipostatic regulation. However, in Paper I, low and high P/L ratio was fed throughout the trial and it is therefore not possible to isolate the effect of body fat accumulation. To be able to elucidate this, the groups should have been fed the same diet from July to September. The isolated effect of body fat levels was tested in Paper II. In Paper II, low, high P/L ratios, and restricted ration (~ 50%) of the diet with high P/L ratio, were used to alter the lipid deposition (primarily muscle fat content) prior to the autumn in large salmon. In this study, clear treatment effect on body fat was observed after three months (May - August). The salmon fed low P/L ratio had higher fat content than those fed high P/L ratio, both in muscle (16.4 vs. 13.2 %) and viscera (39 vs. 29 %).

Restricting the ration of the high P/L diet to 50 % further reduced the fat content to 11.3

% in muscle and 27 % in the viscera. PIT-tagged individuals from the groups with different lipid content were restocked and mixed in the same pens, and then fed the same diet for seven months (August – March). The results showed that the the lipid content in August was negatively correlated to the subsequent TGC and weight gain from August to October. In other words, the weight gain was the highest for the restricted group, intermediate for the high P/L ratio and the lowest for the low P/L ratio group. In October, after two months of feeding a common diet, the muscle fat content was similar in fish from all three groups, whereas the differences in visceral fat content disappeared after four months (December). In the commercial scale trial (Paper IV), no differences in fat content or growth between the groups fed high and low P/L ratio were

(34)

detected prior to or during the autumn. This could be related to lower temperatures in spring and early summer prior to the autumn compared to Paper I and II. Differences in fish size and dietary composition can also be influencing factors. The main increase in fat content among the group fed high P/L ratio in Paper IV took place in December. This late dietary change seemed to have negative consequences for health-related parameters, which are discussed in section 7.5. The results from Paper I and II show that alterations in body lipid levels prior to falling day length in the autumn, significantly affected feed intake and growth in salmon. The results also demonstrate that salmon is able to replenish lipid stores rapidly after dietary lipid restriction and that energy intake and storage is of high priority. The increased growth and rapid replenishment of lipid stores suggest the existence of a robust mechanism for the regulation of body fat in salmonids, and are in line with the previous observation from Silverstein et al. (1999).

7.2 Compensatory growth

The strength of compensatory growth responses seem to depend on the reduction in body condition, length and mass in the restricted or deprived fish groups compared to their fully fed conspecifics (Alvarez and Nicieza, 2005; Johansen et al., 2001; Johnsson and Bohlin, 2006, 2005). In Paper II, feeding a high P/L feed at full or restricted rations resulted in slightly and markedly lower body weight, respectively, compared to feeding the low P/L feed from May to August. Thus, the subsequent increased subsequent growth in fish from these groups from August to October, when all fish were fed a common diet, may partly be explained by the differences in mass and length compared to the low P/L group. However, the small difference in weight between the high and low P/L group in August and the strong correlation between body fat and growth indicate that fat content (i.e. energy status) seem to be a key growth regulator from August to October. This is in line with the observation in Paper I. In the the two last periods of the trial in Paper II, the high P/L ratio and restricted group showed a significantly increased growth compared to the low P/L-ratio group. The high P/L ratio group had significantly higher final body weight than the low P/L ratio group, whereas the restricted group ended up with a similar final body weight as the low P/L ratio group. There are evidences indicating that compensatory growth responses will cease as lipid stores and body condition are restored to levels similar to the unrestricted conspecifics (Ali et al.,

(35)

2003; Alvarez and Nicieza, 2005; Johansen et al., 2001; Johnsson and Bohlin, 2005).

However, the increased growth in the two last periods of the trial in Paper II, from October to March, was evident although the differences in body weight, length and lipid content between the groups had been offset, except for a small difference in length between the restricted and the low P/L ratio group. This apparent overcompensation may indicate the existence of a more complex explanation than just lipostatic regulation.

The growth responses found in Paper II, are somewhat similar to compensatory growth responses and lipostatic regulation observed in previous studies (Ali et al., 2003; Jobling and Johansen, 1999; Johansen et al., 2002, 2001). However, some of the novelty of Paper II is that alterations in body fat content were dietary induced prior to the autumn. Thus, it differs from many previous experiments related to compensatory growth and lipostatic regulation in the exposure of the fish to ambient temperature and day length, which are rigid environmental cues in fish. The autumn is a period associated with optimal temperatures, declining day length, increased feed intake and high lipid deposition. In Paper II, it is discussed how the strong growth response shown by the high P/L ratio and the restricted group may have been triggered by the reproductive life strategy of the Atlantic salmon. The marked increase in lipid deposition and weight gain of the groups fed with a low fat diet may be a response to deposit and store energy for the upcoming spring period, which seems to be a critical period for the initiation and proceeding of the maturation process. Too low energy and fat levels may arrest the maturation process and postpone reproduction (Duston and Saunders, 1999; Rowe et al., 1991; Rowe and Thorpe, 1990; Thorpe, 1994; Thorpe et al., 1990). However, to verify this, the groups of salmon needs to be studied for a longer period and measurements of relevant plasma hormones, gonad development and transcript abundance of relevant genes should be conducted. Although the muscle and visceral cavity are the main sites for lipid deposition in salmon, lipids are also allocated to other parts of the body such as head and bones (Jobling et al., 2002). These structural compartments can also be affected by dietary lipid level and it has been suggested that the lean body mass relative to fat can be important for lipostatic regulation (Jobling and Johansen, 1999; Johansen et al., 2003). The indexes or the fat content of such structural components were not measured in Paper II. However, it should be noted that the visceral fat content was consistently lower in the high P/L ratio and in the restricted group compared to the low

(36)

P/L ratio group throughout the trial, but being significantly lower only in October. In addition, the high P/L and restricted groups had lower dispersion/variation in final body weights and shape compared to the low P/L group. This can indicate that several individuals in the low P/L group showed lowered or impaired growth in latter stages of the trial. These factors may also be a contributing to an increased growth and the obtained weight differences. Even though it could be difficult to pinpoint the causes of the observed growth response, the results in Paper I and II show that increasing the dietary P/L ratio prior to the autumn increase growth and weight gain in salmon.

7.3 Dietary protein-to-lipid ratio and growth

In Paper I, unlike Paper II, low and high P/L ratio were fed throughout the trial and it is therefore not possible to isolate the direct effect of diet on growth from the indirect effect caused by different body fat accumulation. Einen and Roem (1997) evaluated different dietary DP/DE-ratios for Atlantic salmon in seawater and concluded that salmon grown from 1 to 3 kg required a DP/DE above 16.4 g MJ kg^-1 (P/L-ratio of 1.23) for optimal growth performance. Although using somewhat smaller fish, the growth data in Paper I are in agreement with these results. Hence, the low P/L ratio group was fed a dietary DP/DE ratio of 15.2 (P/L ratio of 1.12), whereas the high P/L-ratio group was fed a DP/DE ratio of 17.4 (P/L ratio of 1.40) from July to September. However, in the study of Einen and Roem (1997) diets with different energy content were tested, whereas in Paper I isoenergetic diets were used. Salmonids seem to adjust their appetite according to the dietary energy level (Bendiksen et al., 2002), and it has been suggested that this could be the predominant factor for feed intake and utilization. Studies using isoenergetic diets, or adjusting the dietary ration level so that the diets tested were fed isoenergetically, found no negative effects of reduced dietary P/L ratio on growth performance in Atlantic salmon (Azevedo et al., 2004; Hillestad et al., 1998; Karalazos et al., 2011, 2007). These finding differ from the results from Paper I. In the experiments of Karalazos et al. (2011, 2007), isoenergetic diets with a P/L ratio in the range of 0.75 – 1.19 (DP/DE ratio: 12.5 - 15.5 g MJ kg^-1) and low fishmeal inclusion were used, whereas Azevedo et al. (2004) tested the diets using a wild salmon strain reared in freshwater with a constant temperature of 8 °C. Thus, the mentioned studies have clear dissimilarities in dietary inputs and experimental design compared to those from Paper

(37)

I. In addition, the salmon used were smaller (0.1 – 1 kg) than the salmon used in the studies of Karalazos et al. (2011, 2007). In Paper I, the seawater temperature from July to September had a mean of 13.6ºC, which is associated with optimal growth rates for salmon post-smolt (Handeland et al., 2008). Small salmonids require higher dietary proportions of digestible protein than larger salmonids (Cho and Kaushik, 1990; Einen and Roem, 1997), and this seems to be particularly important at high temperatures and during rapid somatic growth (Bendiksen et al., 2003). These mentioned factors, together with the potential lipostatic regulation, may explain why the results obtained in previous studies using isoenergetic diets differ from those presented in Paper I. The results also suggest that a DP/DE of 15.2 g MJ kg^-1 (P/L-ratio of 1.12) is below the required P/L ratio to attain optimal/maximum growth of S1 salmon in the weight segment from 0.1 – 1 kg.

7.4 Nutrient utilization

FCR and nutrient retention are commonly used to evaluate the efficiency and sustainability of aquaculture systems and diets. For an optimal and sustainable production, low FCR and high nutrient retention are favorable. High dietary energy content seems important for maintaining low FCR when fish size increases (Einen and Roem, 1997; Hillestad et al., 1998). In Paper I, no significant differences in FCR were detected between the low and high dietary P/L groups. However, when the FCR was calculated based on gutted weight (FCRg), the low P/L ratio group had a significantly higher FCRg from July to September compared to the high P/L ratio group. While visceral somatic index (VSI) of the group fed the high dietary P/L ratio was relatively stable during the experiment, the VSI of the group fed the low dietary P/L ratio increased gradually. Consequently, the low P/L group had significantly higher VSI and lower carcass yield compared to the high P/L group at the end of the trial. In Paper II, feeding the high P/L feed or a restricted ration of this diet from May to August also significantly reduced the VSI and reduced the visceral fat content compared to feeding a low P/L ratio. Refstie et al. (2001) found that feeding high-fat diets resulted in increased visceral fat content, lower carcass yield and larger visible fat deposits. In their study, the high-fat diets increased the weight gain by 122 grams compared to medium-fat diets;

however, 91 grams of this weight gain were lipid. This and the results from Paper I and II shows that a major part of the fat accumulated in salmon when feeding high-fat diets

(38)

ends up around the intestines and in other fat deposits, which are often removed by gutting and trimming during slaughter. Hence, high dietary fat content may increase adiposity and do not primarily enlarge the carcass weight, that is the main edible product for sale and holds the most value. Although, it is possible to utilize the viscera for oil extraction and/or native protein and hydrolysates (Villamil et al., 2017; Wu and Bechtel, 2008), salmon feeding strategies should be optimized to maximize the carcass production.

The nutrient retention gives a measure of the proportion of the dietary nutrient that is retained in the fish. Simplified, there is generally an inverse relationship between the dietary inclusion rate of protein and lipid, and the retention efficiency of these nutrient in salmon (Bendiksen et al., 2003; Einen and Roem, 1997; Hillestad et al., 1998). Thus, reducing the P/L ratio can increase the retention efficiency of protein and facilitate a favorable “protein sparing” effect (Azevedo et al., 2004; Hillestad et al., 1998; Johansen et al., 2003; Karalazos et al., 2011, 2007), whereas the retention of lipid can potentially decrease to some extent (Weihe et al., 2018). It should be noted that lipids, unlike energy and protein (as N x 6.25) can be synthesized de novo by the fish. The term “lipid retention” should, therefore be used with care. However, it is a useful tool to express dietary or seasonal variations in energy deposition. In Paper I, the apparent nutrient retention efficiency of protein, lipid and energy was measured based on whole body composition (referred to as relative nutrient retention). The releative retention of lipid was low from April to June (25-28%), intermediate from June to July (45-50%) and high from July to September (67-75%). Generally, the lipid and energy retention increased with increasing feed intake, growth, weight gain and dietary energy level. These seasonal patterns are consistent with the observations from Alne et al. (2011). In their study, S1 post-smolt had lower lipid retention during the spring (~20%) compared to the autumn (~60%) and this coincided with low and high growth rates, respectively. In general, energy retention increases with increased feed intake and growth, because the relative proportion of the eaten energy used for maintenance is reduced, as shown for salmon by Helland et al. (2010). The relationship between growth and retention is even more marked for lipids, since fish eating at a high rate will deposit surplus energy as body fat stores. Accordingly in Paper I, the retention of protein was reasonably stable (at approximately 50%) and less dynamic than the retention of lipid, as previously reported

(39)

(Alne et al., 2011; Nordgarden et al., 2003b). This can be explained by the relatively stable body content of protein compared to that of lipid, which vary largely with energy supply in salmonids. This is also shown by assessing the whole body nutrient composition of the dietary groups during the experiment described in Paper I.

The nutrient retention efficiency (referred to as in Paper I as relative nutrient retention) expresses utilization efficiency and do not show how much of the nutrient that is deposited in the fish from the diet, when diets that differ in protein and lipid content are compared. To illustrate this in a simple manner, the retention of the nutrients from the feed were calculated in Paper I and expressed in grams (referred to as absolute nutrient retention). The increase in P/L ratio was not synonymous with reduced retention of protein in Paper I. No significant differences in relative protein retention were detected between the dietary treatments during the initial and latter stages of the trial. However, a significantly lower protein retention among the high P/L ratio group compared to the low P/L-ratio group was detected from June to July. In the periods when there were no differences in relative protein retention, the fish fed high the P/L ratio had significantly increased absolute protein retention. Thus, the fish fed the high P/L ratio retained more protein from the diet compared to the fish fed the low P/L ratio. This indicates that the increased protein content in the high P/L ratio diets was efficiently utilized for growth and weight gain. In the last period of the trial, the increased absolute protein retention in the high P/L group could be explained by a combination of increased whole body protein content, significantly improved feed intake, growth and weight gain compared to the low P/L ratio group. This was also accompanied by higher condition factor and carcass yield. The relative and absolute retention of lipid was lower in the high compared to the low P/L ratio group in the period from June to July. This was reflected in a significantly lower whole body lipid and energy content within the group fed the high compared to the low P/L ratio in the end of July. Hence, a large part of the dietary lipid content was used for energy production and less for storage in the high P/L ratio group. The period from June to July has previously been identified as a period when S1 salmon smolt have low feed intake, growth and lipid retention (Alne et al., 2011; Rørvik et al., 2007). From July to September, a significantly higher lipid retention was observed for the high compared with the low P/L ratio group. However, no significant differences were detected in absolute lipid retention during the period from July to September.

Growth, feed utilization, health and biometric parameters in Atlantic salmon (Salmo salar L.) - Influence of dietary protein-to-lipid ratio and body fat status

Jens-Erik Dessen

Growth, feed utilization, health and biometric

parameters in Atlantic salmon (Salmo salar L.) - Influence of dietary protein-to-lipid ratio and body fat status

Vekst, fôrutnyttelse, helse og biometriske parameter hos Atlantisk laks (Salmo salar L.) – Effekt av protein-til-fett forholdet i fôret og fettinnholdet i laksen

Growth, feed utilization, health and biometric parameters in Atlantic salmon ( Salmo salar L.) - Influence of dietary

protein-to-lipid ratio and body fat status

Acknowledgements

Table of contents

1. Abbreviations

2. List of publications

3. Summary

Sammendrag

4. General introduction 4.1 Atlantic salmon aquaculture

4.2 Feed for farmed Atlantic salmon

4.3 Diseases in salmon aquaculture

4.4 Growth and feed utilization

4.5 Seasonal variations in growth and lipid deposition

4.6 Fat storage as a reproductive adaption?

4.7 Compensatory growth and lipostatic regulation

5. Objectives

e nt al o v er v ie w

7. Main results and general discussion

7.1 Body fat levels and growth

7.2 Compensatory growth

7.3 Dietary protein-to-lipid ratio and growth

7.4 Nutrient utilization